Hard Drive as slave keeps system from booting

[jack]

Have a Western Digital that stopped booting. It hangs at the Windows
XP spash screen, with the scroll bar running beneath it. When trying
to boot to safe mode, it hangs as it's loading .sys files from the
drivers folder (last one listed is amdagp.sys). I can see it in the
bios, and I've run a WD diagnostic tool and another one on it and they
say it's fine. I'm guessing the MBR might need fixing.

But the real problem is I put it in another system as a slave and it
keeps that one from booting up, too. Same problem with getting to safe
mode and a prompt, except that it hangs at agpcpq.sys.

Any thoughts, on at least being able to access it as a slave from a
prompt so I could try to repair it, or at least save the data? TIA!

 

[kony]

Have a Western Digital that stopped booting. It hangs at the Windows
XP spash screen, with the scroll bar running beneath it. When trying
to boot to safe mode, it hangs as it's loading .sys files from the
drivers folder (last one listed is amdagp.sys). I can see it in the
bios, and I've run a WD diagnostic tool and another one on it and they
say it's fine. I'm guessing the MBR might need fixing.

No, not possible.

If the MBR needed fixed it couldn't begin to boot the OS.
Since it does this much it would be something else, most
likely a driver or OS problem.

How long had system worked properly prior to this?
What was the last change(s) made to system, since the last
time it had successfully booted?

You could try a repair install of XP, or hook up another
drive, do clean install of XP and THEN add the WD drive back
to copy off data. There's a fair chance nothing is wrong
with the drive so you might then retry it gradually while
relying on the other drive.

But the real problem is I put it in another system as a slave and it
keeps that one from booting up, too. Same problem with getting to safe
mode and a prompt, except that it hangs at agpcpq.sys.

Is the other drive a WD? I'm wondering if you have jumper
problems. Seems very odd that it's stopping at AGP drivers
both times, I wonder what loads next in the boot sequence.
You might Google for a boot log and check that angle.

Any thoughts, on at least being able to access it as a slave from a
prompt so I could try to repair it, or at least save the data? TIA!

From a prompt? Is it NTFS?
How about if you dont' install it as slave in the other
system, rather Secondary Master (alone on that IDE cable,
jumpered as "single" drive).

If all else fails you might try hooking it up to a drive
enclosure (one tested as working per the system used, with
some other drive) such that the OS is done loading and you
can see if it appears. Naturally the enclosure would have
to support the capacity of the drive, some have limits like
128 or 160GB, my memory of specifics is a bit vague on this.

 

[MF]

there's no form of windows corruption i ever hard of that will stop a system
from booting when the windows installation is on the slave, and the master
is the boot drive. is the drive spinning? i've seen wd diag programs give
good results on a shorted out drive one time, and then fail to access the
drive the next five times.

virus? reinstall the driver. but this seems unlikely, if, after
disconnecting it (as slave) the system will boot, then fail to boot after
reconnecting it as slave.

fix the mbr, what the hell, what have you got to lose. btw, this is a
symptom of a shorted out component - but the boot to the xp splash screen
doesn't fit that scenario and you diag of where it stops doesn't fit that
scenario.

download knoppix linux, make the cd, and see if the system will boot off the
knoppix cd, then try accessing the drive. if you can access it , you can
get the data. then try copy over that amdagp driver you mentioned with a
working version and try it out. no joy, then wipe the whole thing and
reinstall.

also, describe in more detail what happens when you boot with that drive as
a slave- you're not trying to boot from it, right?

good luck. this sounds like a true PITA.

 

[jack]

No, not possible.

If the MBR needed fixed it couldn't begin to boot the OS.
Since it does this much it would be something else, most
likely a driver or OS problem.

How long had system worked properly prior to this?
What was the last change(s) made to system, since the last
time it had successfully booted?

In its current configuration, not sure. Same with not knowing what
changes had been since last reboot, as I leave it on awhile. Nothing
hardware-wise, though.
I should've mentioned this before, but before I shut it down prior to
it not booting, I had experienced low memory warnings, which prompted
me to close everything out and reboot. Any thoughts on what might've
happened? I tested the memory and the MS Memory Diagnostics disk said
it's fine.

You could try a repair install of XP, or hook up another
drive, do clean install of XP and THEN add the WD drive back
to copy off data. There's a fair chance nothing is wrong
with the drive so you might then retry it gradually while
relying on the other drive.

Can't do a repair (Recovery Console) as it hangs at "examining ___ MB
on Disk (0) at ...."
I suspect I'd have the same problem trying to slave this to a new
drive in the same system, as trying to slave it on another system. But
as a last resort I'll try it.

Is the other drive a WD? I'm wondering if you have jumper
problems. Seems very odd that it's stopping at AGP drivers
both times, I wonder what loads next in the boot sequence.
You might Google for a boot log and check that angle.

It's not a WD. I tried it as both Cable Select and Master Slave, and
also, I put the second drive (slave) from the dead system in the other
system and it booted fine. So it is tied to this drive.
I was wondering what was next in the boot sequence, too. Didn't know
if the driver it shows is the problem, or the next one. But also, the
other system where it won't boot as a slave, is not an AMD system?!
Why would it be trying to load AMD drivers? Where does the command to
load the drivers come from?
I found the following supposed boot sequence on the web:
"Ntldr reads the registry files, selects a hardware profile,
control set and loads device drivers, in that order.
Then, Ntoskrnl.exe takes over and starts Winlogon.exe which starts
Lsass.exe (Local Security Administration), this is the program
that displays the Welcome screen (If Professional Edition-the
Windows Log On dialog box), and allows the user to log on
with his/her user name and password."
If it's accurate, could I be getting to the WinXP splash screen if all
the drivers didn't load properly? I only see the drivers loading, and
stopping, when I try to boot to Safe Mode. Is the sequence, and
drivers, different for Safe Mode? Or could it be moving on to the
splash screen eventhough the drivers are still loading in the
background?
Also, the second, working system, when the drive is slaved to it,
halts when trying to get to Safe Mode at the entry following
amdagp.sys, which is agpCPQ.sys. And, fwiw, the first time that I did
it, I thought I saw a quick flash at amdagp.sys to hit (ESC?) to skip
the file, but then it continued on and halted at the next entry. And I
no longer see a prompt.
Searching my registry in the working system found entries for these
files at:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\amdagp
Display Name REG_SZ AMD AGP Bus Filter Driver
ImagePath REG_EXPAND_SZ System32\Drivers\amdagp.sys
And for agpcpq.sys,
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\agpCPQ
but entitled Compaq AGP Bus Filter.
They're also both in ControlSet001 and ControlSet003.

Do I dare delete these values/entries? Of course, I still couldn't do
it on the problem machine. But how are these tied to a slave drive?!

However, something interesting I just found in a thread at a forum.
The gist of it was dealing with hangs at mup.sys, but one guy got past
that, and the next one was agp.sys, where it still hung. But that
thread had a few different fixes, apparently because of a few
different causes. If anyone's interested, and think they could help
use the info to figure it out, it's at:
http://www.annoyances.org/exec/forum/winxp/t1047532372

From a prompt? Is it NTFS?
How about if you dont' install it as slave in the other
system, rather Secondary Master (alone on that IDE cable,
jumpered as "single" drive).

Yes, it's NTFS. I'll try that, though see below re: enclosure, which
probably makes this moot.

If all else fails you might try hooking it up to a drive
enclosure (one tested as working per the system used, with
some other drive) such that the OS is done loading and you
can see if it appears. Naturally the enclosure would have
to support the capacity of the drive, some have limits like
128 or 160GB, my memory of specifics is a bit vague on this.

I got one, and the slave from the dead system got recognized right
away in my working system, with drive letters showing up in My
Computer.
When I hooked up the "dead" drive (which I actually did first), there
was a pop-up in the SysTray indicating a new device, and it did show
it as WD 6GB, but no drives showed up in My Computer, and the system
slowed to a MAJOR crawl. Couldn't get Task Manager up and had to
ultimately re-boot. Tried a few times, with the same, except nothing
in the SysTray after the initial hookup.

Any more thoughts? Help!

 

[kony]

In its current configuration, not sure. Same with not knowing what
changes had been since last reboot, as I leave it on awhile. Nothing
hardware-wise, though.
I should've mentioned this before, but before I shut it down prior to
it not booting, I had experienced low memory warnings, which prompted
me to close everything out and reboot. Any thoughts on what might've
happened? I tested the memory and the MS Memory Diagnostics disk said
it's fine.

Try something better than the MS Memory Diagnostics-
Memtest86 or Memtest86+ (Google for them) ran for several
hours.

There is insufficient evidence to draw a conclusion but I
suspect you're infeced with some sort of virus, malware,
spyware or similar, which initially caused the problem and
is now crashing windows regardless of whether that was the
code's "original", intended purpose. If you have antivirus
boot disks and can scan the drive(s) you might try that.

Can't do a repair (Recovery Console) as it hangs at "examining ___ MB
on Disk (0) at ...."
I suspect I'd have the same problem trying to slave this to a new
drive in the same system, as trying to slave it on another system. But
as a last resort I'll try it.

Well now I wonder if it's coincidence that this is occurring
alongside the windows out-of-memory messages, as these two
quite different symptoms are not usually by a (single)
common cause.

When slaving the drive be clear on the jumper settings for
that drive AND the other, existing master. Other than that
it would seem likely to have a hang in current system if it
caused this in the other. This type of problem should be
caught by the drive manufacturer's diagnostics, did you
(re)run that?

It's not a WD. I tried it as both Cable Select and Master Slave, and
also, I put the second drive (slave) from the dead system in the other
system and it booted fine. So it is tied to this drive.
I was wondering what was next in the boot sequence, too. Didn't know
if the driver it shows is the problem, or the next one. But also, the
other system where it won't boot as a slave, is not an AMD system?!

It might not be that driver OR a next one, could be
something else windows it doing afterwards. Try going to
the repair console and running command;
C:\> chkdsk C: /R

Why would it be trying to load AMD drivers? Where does the command to
load the drivers come from?
I found the following supposed boot sequence on the web:
"Ntldr reads the registry files, selects a hardware profile,
control set and loads device drivers, in that order.
Then, Ntoskrnl.exe takes over and starts Winlogon.exe which starts
Lsass.exe (Local Security Administration), this is the program
that displays the Welcome screen (If Professional Edition-the
Windows Log On dialog box), and allows the user to log on
with his/her user name and password."
If it's accurate, could I be getting to the WinXP splash screen if all
the drivers didn't load properly? I only see the drivers loading, and
stopping, when I try to boot to Safe Mode. Is the sequence, and
drivers, different for Safe Mode? Or could it be moving on to the
splash screen eventhough the drivers are still loading in the
background?
Also, the second, working system, when the drive is slaved to it,
halts when trying to get to Safe Mode at the entry following
amdagp.sys, which is agpCPQ.sys. And, fwiw, the first time that I did
it, I thought I saw a quick flash at amdagp.sys to hit (ESC?) to skip
the file, but then it continued on and halted at the next entry. And I
no longer see a prompt.
Searching my registry in the working system found entries for these
files at:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\amdagp
Display Name REG_SZ AMD AGP Bus Filter Driver
ImagePath REG_EXPAND_SZ System32\Drivers\amdagp.sys
And for agpcpq.sys,
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\agpCPQ
but entitled Compaq AGP Bus Filter.
They're also both in ControlSet001 and ControlSet003.

Do I dare delete these values/entries? Of course, I still couldn't do
it on the problem machine. But how are these tied to a slave drive?!

Since the system previously worked and you hadn't JUST
finished installing a new driver, I doubt the solution is
deleting any of those files... That should be how it was
the last time it'd booted successfully so there is something
more at play here. Moving the drive to the other system
should indicate that the drive is the problem, unless of
course these two boxes were networked and a virus had spread
to the 2nd system, it was merely that the 2nd system hadn't
been rebooted yet when you did so to try this hard drive in
it... just throwing ideas out there, grab whatever applies.

However, something interesting I just found in a thread at a forum.
The gist of it was dealing with hangs at mup.sys, but one guy got past
that, and the next one was agp.sys, where it still hung. But that
thread had a few different fixes, apparently because of a few
different causes. If anyone's interested, and think they could help
use the info to figure it out, it's at:
http://www.annoyances.org/exec/forum/winxp/t1047532372

If the drive still checks out ok with the manufacturer's
diagnostics I'd see about copying off data somehow (maybe
the USB enclosure idea I mentioned previously) or as master
alone on the secondardy ide channel. You could try a repair
install again or just use another HDD. If the prior
suggestions don't let you access the drive then I hope you
made backups of anything important.

 

[jack]

Try something better than the MS Memory Diagnostics-
Memtest86 or Memtest86+ (Google for them) ran for several
hours.

I'll do that; thanks.

There is insufficient evidence to draw a conclusion but I
suspect you're infeced with some sort of virus, malware,
spyware or similar, which initially caused the problem and
is now crashing windows regardless of whether that was the
code's "original", intended purpose. If you have antivirus
boot disks and can scan the drive(s) you might try that.

Guess it's possible. I've now put a new drive and re-installed XP on
my "dead" system, loaded AVG on and in doing so, created rescue boot
discs with AVG. I'll swap out the new drive with the bad one and try
the discs. I so often come across errors when doing stuff that
requires me to either shut off my pop-up blocker/firewall/anti-virus
to allow what i'm trying to do to work, that I may have and forgotten
to re-enable it. But I've had Kaspersky active non-stop on the other
machine, that is indeed networked, and don't recall any hits on it.
Though that may not necessarily have happened, I guess.

Well now I wonder if it's coincidence that this is occurring
alongside the windows out-of-memory messages, as these two
quite different symptoms are not usually by a (single)
common cause.

I wasn't sure if, when running out-of-memory, it caused errors on the
hard drive, such as corrupted drivers. I don't know how out-of-memory
issues manifest in the system.

When slaving the drive be clear on the jumper settings for
that drive AND the other, existing master. Other than that
it would seem likely to have a hang in current system if it
caused this in the other. This type of problem should be
caught by the drive manufacturer's diagnostics, did you
(re)run that?

Yeah, I ran the WD Diagnostic Tool, as well as one that's on the
Ultimate Boot Disk, TestDisk, if you're familiar with that. Both say
they didn't find any errors. The author of TestDisk said to save the
log of the test, though, and email it to him. Unfortunately, it was
saved to the drive that I can't access.

It might not be that driver OR a next one, could be
something else windows it doing afterwards. Try going to
the repair console and running command;
C:\> chkdsk C: /R

No can do; Recovery Console stops at "examining 588000 or whatever MB
of disk (0) on ... I've now found that that only splashes for a
second prior to the Recovery Console starting. Nevertheless, the
console never starts to the point where I can do anything in it.

Since the system previously worked and you hadn't JUST
finished installing a new driver, I doubt the solution is
deleting any of those files... That should be how it was
the last time it'd booted successfully so there is something
more at play here. Moving the drive to the other system
should indicate that the drive is the problem, unless of
course these two boxes were networked and a virus had spread
to the 2nd system, it was merely that the 2nd system hadn't
been rebooted yet when you did so to try this hard drive in
it... just throwing ideas out there, grab whatever applies.

Except that the second system only has problems booting while the bad
drive is slaved to it; take it off and it's fine. So don't think it
would be a virus.

If the drive still checks out ok with the manufacturer's
diagnostics I'd see about copying off data somehow (maybe
the USB enclosure idea I mentioned previously) or as master
alone on the secondardy ide channel. You could try a repair
install again or just use another HDD. If the prior
suggestions don't let you access the drive then I hope you
made backups of anything important.

Sorry, my last post was so long that the part where I mentioned the
enclosure must've gotten snipped.
I got one, and the slave from the dead system got recognized right
away in my working system, with drive letters showing up in My
Computer.
When I hooked up the "dead" drive (which I actually did first), there
was a pop-up in the SysTray indicating a new device, and it did show
it as WD 6GB, but no drives showed up in My Computer, and the system
slowed to a MAJOR crawl. Couldn't get Task Manager up and had to
ultimately re-boot. Tried a few times, with the same, except nothing
in the SysTray after the initial hookup.

Thanks again for all your efforts. Much appreciated!

 

[jack]

I'll do that; thanks.

Guess it's possible. I've now put a new drive and re-installed XP on
my "dead" system, loaded AVG on and in doing so, created rescue boot
discs with AVG. I'll swap out the new drive with the bad one and try
the discs. I so often come across errors when doing stuff that
requires me to either shut off my pop-up blocker/firewall/anti-virus
to allow what i'm trying to do to work, that I may have and forgotten
to re-enable it. But I've had Kaspersky active non-stop on the other
machine, that is indeed networked, and don't recall any hits on it.
Though that may not necessarily have happened, I guess.

I wasn't sure if, when running out-of-memory, it caused errors on the
hard drive, such as corrupted drivers. I don't know how out-of-memory
issues manifest in the system.
Yeah, I ran the WD Diagnostic Tool, as well as one that's on the
Ultimate Boot Disk, TestDisk, if you're familiar with that. Both say
they didn't find any errors. The author of TestDisk said to save the
log of the test, though, and email it to him. Unfortunately, it was
saved to the drive that I can't access.
No can do; Recovery Console stops at "examining 588000 or whatever MB
of disk (0) on ... I've now found that that only splashes for a
second prior to the Recovery Console starting. Nevertheless, the
console never starts to the point where I can do anything in it.
Except that the second system only has problems booting while the bad
drive is slaved to it; take it off and it's fine. So don't think it
would be a virus.

Sorry, my last post was so long that the part where I mentioned the
enclosure must've gotten snipped.
I got one, and the slave from the dead system got recognized right
away in my working system, with drive letters showing up in My
Computer.
When I hooked up the "dead" drive (which I actually did first), there
was a pop-up in the SysTray indicating a new device, and it did show
it as WD 6GB, but no drives showed up in My Computer, and the system
slowed to a MAJOR crawl. Couldn't get Task Manager up and had to
ultimately re-boot. Tried a few times, with the same, except nothing
in the SysTray after the initial hookup.

Thanks again for all your efforts. Much appreciated!

Well, lo and behold, I've got it fixed. Now if I only knew how. I
installed XP Pro on a new drive, in the old dead system. Everything
worked fine, and decided to plug in the dead drive in the enclosure,
and whadya know, it was recognized! Didn't change anything on it from
when I last tried it on the other machine which would lock up/slow WAY
down. The differences between the systems, which I don't think account
for it:
Old, formerly dead, System: Athlon 850 MHz, 512 MB Ram, new install of
XP SP2 incorporated
Other System, where it would lock up: Celeron 2.6 GHz, 512 Ram, XP
Home no SP2

Now, for the details of what was wrong. As I said, it was finally
recognized, but it said that the M: drive (former C:/boot drive) was
corrupted. But now I could access it from a prompt, so I ran chkdsk /r
I'll paste the results of it below. Anyone have any thoughts as to
what happened to cause the below-indicated problems? Thanks again for
any and all help you provided!
Oh yeah, one more thing. I'm not able to access some folders of the
drive, I'm sure because I'm not logged into it. Is there a way to
change the permissions/share the files, without using it to boot up a
system? Thanks.

C:\chkdsk /r m:
The type of the file system is NTFS.

CHKDSK is verifying files (stage 1 of 5)...
Deleted corrupt attribute list entry
with type code 128 in file 31110.
Deleting corrupt attribute record (128, $J)
from file record segment 25635.
Deleting corrupt attribute record (128, $J)
from file record segment 47593.
File verification completed.
Deleting orphan file record segment 25635.
CHKDSK is verifying indexes (stage 2 of 5)...
Index verification completed.
CHKDSK is recovering lost files.
Recovering orphaned file ZA2482~3.TXT (10355) into directory file
33901.
Recovering orphaned file ZALog2005.02.10.txt (10355) into directory
file 33901.
Correcting a minor error in file 31110.
Inserting an index entry into index $I30 of file 11.
CHKDSK is verifying security descriptors (stage 3 of 5)...
Security descriptor verification completed.
CHKDSK is verifying Usn Journal...
Creating Usn Journal $J data stream
Usn Journal verification completed.
CHKDSK is verifying file data (stage 4 of 5)...
File data verification completed.
CHKDSK is verifying free space (stage 5 of 5)...
Free space verification is complete.
CHKDSK discovered free space marked as allocated in the
master file table (MFT) bitmap.
CHKDSK discovered free space marked as allocated in the volume bitmap.
Windows has made corrections to the file system.

8193118 KB total disk space.
7859344 KB in 62270 files.
23720 KB in 4267 indexes.
0 KB in bad sectors.
143458 KB in use by the system.
43024 KB occupied by the log file.
166596 KB available on disk.

4096 bytes in each allocation unit.
2048279 total allocation units on disk.
41649 allocation units available on disk.